Quote: |
So when I saw that, I started thinking about how to do a cardioid sub pattern to reduce that energy on the stage. And with the help of fellow labster Harry Brill through IMs and transferring MAPP files back and forth, and other useful insight from Phill Graham and David Buehler, we ended up with the following: |
gtphill wrote on Mon, 03 January 2005 20:23 |
This technique also works in other configurations. In your narrow "shoebox" proscenium soft-seater, you can do a horizontal line array of subwoofers across the front of the stage area to eliminate power alley, and then do a second line of subs, like mike has done, to keep the energy from spilling off the back of the subs onto the stage. |
Jason wrote on Mon, 03 January 2005 16:50 |
Hi Mike, Great case study. I did an almost identical show on new year's eve, arrayed like your "bad" example. My mon beach was right in that dark red spot behind a corner array... talk about getting your bell rung. There's nothing like spending the first 10 seconds of 2005 chasing down RF's so that they don't drive off the rink in the rappers' Mercedes Have you done any of the same modeling for standard ampitheatre / end stage setups? I wonder if that technique would still be worthwhile when the performers aren't directly behind the subs. -Jason |
Tom Reid wrote on Tue, 04 January 2005 21:25 |
I'm following this thread, and I think I've absorbed about %90 of it. Can I ask, when you fly subs, what do you need to deal with as far as the "Z" axis? How does cardiod work "up there"? |
gtphill wrote on Mon, 03 January 2005 22:23 |
For those unfamiliar with the "cardiod sub" technique, let me explain what is done processing-wise. The rear row of subwoofers is delayed to line up with the spill from the front ones. The rear boxes have inverted polarity. The result is that the waves cancel in behind the two rows of boxes, and sum coherently in front. The wavelength of maximum cancellation is a function of the front-to-back driver spacing. The longer the spacing, the lower the frequency. |
Eytan Gidron wrote on Wed, 05 January 2005 09:02 |
Thanks Phill for this explenation. I am not sure that I got it right. Can you get a bit more specific - can you give an example with actual numbers (such as which delay time to which spacing, which frequencies will it affect?) Will this method work well with horn loaded subs (such as the Turbo subs)? |
gtphill wrote on Mon, 03 January 2005 15:23 |
Mike also could have placed the two subs back to back (i.e. speakon to speakon) and then delayed the front boxes. DF file for pictures of a horizontal sub line array+cardiod model. |
Andy Leviss wrote on Thu, 06 January 2005 01:03 | ||
I'm not sure I'm entirely following this. Why would you then delay the front ones instead of the back? Isn't turning the back box 180 degrees the same as reversing its polarity, once you account for the different spacing between the drivers from having them located on the other side of the box (which is just a factor of the delay setting, right?)? |
Harry wrote on Thu, 06 January 2005 19:22 |
I work for a Meyer house currently. |
Quote: |
They don't seem to have any issues with teaching anyone and everyone. What makes you think they are trying to hide anything? |
Quote: |
How does the M3D-Sub steer without digital delay? We use analog phase networks to control the cardioid behavior of the M3D and the M3D-Sub. |
Quote: |
I have also used the CD18s and CD12s. The CD12 uses EXACTLY this method, with perhaps a bit of frequency response shading to make up for the fact that the rear is facing back. This is not only good for 1 frequency either. The wavelengths are long. It works very well about an octave in each direction, see for yourself. |
Quote: |
(The GEO T uses a similar method with a low pass filter that has a slope that works with the front speakers to add control as the frequency drops.) Back to Meyer. I don't know what the M3D and M3DSub are doing exactly, |
Quote: |
but according to Mauricio it is not advantages to have rear firing drivers, except for the package and repeatability of an all in one box. The full benefits are only realized with all the boxes facing the same direction. This method with only 2 sources is not he best. A better method uses 4 sourced, spaced 1/4 wavelength of the highest frequency of interest, then delay 3 of them back to the one in the back. This method yields about a 23dB difference front to back. |
Quote: |
I find your attitude disturbing. I would be happy to pass Mauricio's phone number to you, and you can also speak with Dr. Don Pearson, whome I sat next to in the class. |
Quote: |
Would that be proof enough for you that Meyer is interested in improving the quality of audio in the industry as a whole? |
Quote: |
Okay? I'm sorry if my post was less than clear. I wasn't trying to say anything particularly important. Just an idle thought. WM |
Peter wrote on Fri, 07 January 2005 00:18 |
The impulse response is not perfect. Imagine if the waveform was a single pulse - the top half of a 60hz sine wave. The front box will be producing sound for the equivalent of 180 degrees of time (at 60hz) before anything from the rear box is received. What is received (from the rear box) is inverted and delayed by 180 degrees equivalent of time. So the original sound is the top half of a sine wave, but the received sound is a complete sine wave, the original plus a delayed and inverted original. I guess you could call it distortion, it will still sound “nice” and could well be the best compromise in many situations. At back however every thing cancels perfectly. Peter |
Quote: |
Also, for what it's worth, I kind of doubt that Meyer's (and Nexo's, for that matter) cardioid subs use a technique that's this simple, since it works best at only one frequency, and I doubt that Meyer would reveal one of their great secrets to you guys like that anyway. They seem to have figured out how to do this on a broadband basis. |
William Mortensen wrote on Thu, 06 January 2005 20:46 |
Or do you guys have MAPP Online Professional? |
Harry wrote on Wed, 05 January 2005 19:45 |
Speed of sound at 70 Deg F 1129 60Hz center frequency. 1129/60/4 = 4.7 ft (Wavelength (f/cycle) = Speed of Sound / Frequency) Center to center (just measure front of cab to front of cab with both facing forward. 4.7ft is the spacing. 1000/60/4 = 4.17 (Period (ms) = 1000 / Frequency) Set the delay to the rear speaker to 4.17 and reverse the polarity. Keep in mind the speed of sound changes with temperature. |
Harry wrote on Thu, 06 January 2005 19:22 |
I find your attitude disturbing. I would be happy to pass Mauricio's phone number to you, and you can also speak with Dr. Don Pearson, whome I sat next to in the class. Would that be proof enough for you that Meyer is interested in improving the quality of audio in the industry as a whole? |
Dan Mortensen wrote on Sat, 08 January 2005 03:46 |
When the PSW-6 came out, it was the first commercially available cardioid subwoofer that functioned as well as it did. At that time, I spoke with someone within Meyer (not one of the main design engineers, I believe) who explained that the method was not simply reversing polarity and adding delay, because that method would only work at one frequency, and this worked for more than two octaves. Since I don't know any better, that explanation satisfied me. |
Quote: |
If Meyer's method is as complex as they say, and if they are proud of having the first successful commercially available product using what they figured out for the first time, I would not expect them to jump up and offer it free to anyone who would listen. |
Quote: |
That has nothing to do with their interest in improving the quality of audio in the industry as a whole, but, as Will said, with protecting their intellectual property. |
Jon Martin wrote on Mon, 10 January 2005 06:44 |
This definitely has to be one of the best posts in recent memory. |
KeithB wrote on Mon, 10 January 2005 11:50 |
Interesting post! However, in my browser , all the links show with a box with a red X in them and can't be text copied. If I type in the addy this is the result..."Access Denied. Bandwidth limit exceeded." Am I missing something in my browser set-up? |
Langston Holland wrote on Mon, 10 January 2005 08:33 |
Your browsers are fine - I had no idea so many people would read this post - I've just signed up for a "real" ftp storage account and the post should be functioning again within a couple of hours. |
Jon Martin wrote on Mon, 10 January 2005 01:44 |
This definitely has to be one of the best posts in recent memory. A couple of questions: * Deploying this method in a traditional L-R setup, what (possably adverse) inter-reaction would a single center sub and/or lipfills have? ie; would outer lipfills be affected by the polarity flip on the 'rear' subs?... I'm ASSuming spacing would be critical.. |
Quote: |
* Downstage lobes of the sidefills, inter-reaction with those? |
Quote: |
* Are the Nexo CD 12/18 built internally so the elements 'line up' at the exit physically and then some degree of delay and attenuation applied? |
Quote: |
* w/ Subs in the air... Optimal point(s) of measurements? |
Harford Sound wrote on Mon, 10 January 2005 15:37 |
What is MAPP or whatever those graphs are and how do you get the results? Could you explain what everything means on them? Thanks! |
Harford Sound wrote on Mon, 10 January 2005 16:37 |
Man.. What an interesting thread. I started off reading not having a clue what was going on, but now Im starting to get it. I think Im going to have to read it a few more times to truely understand everytinhg. But I have a question. What is MAPP or whatever those graphs are and how do you get the results? Could you explain what everything means on them? Thanks! I feel sorry for the animals in the woods of those pics. That must be a lot of sound. |
Quote: |
Did you actually achieve thorax-rattling in the upper decks? And how bearable was it closer up? |
Dr. Don wrote on Fri, 14 January 2005 23:36 |
Did you actually achieve thorax-rattling in the upper decks? And how bearable was it closer up? Maybe some posters out there who were at some of these shows should answer this |
gtphill wrote on Thu, 06 January 2005 12:06 |
Actually, this a correct way of doing it. Mike B's way: You need to cancel the waves behind the front box. So (for simplicities sake) if the front box is producing cos(x) then the rear box needs to produce -cos(x + D) where D is the time it takes for the sound from the front box to reach the rear. The "-" comes from the polarity invert/box direction, and the D comes by delaying the rear box long enough from the sound from the front box to get there. Alternative way: Space the boxes 1/4 wavelength apart (90 degrees apart). Now delay the front box another 1/4 wavelength (90 degrees). The output from the the front box takes 1/4 wavelength to reach the rear box, and other 1/4 wavelength you added artificially, so cancellation ensues. The undelayed sound from the rear box takes 1/4 wavelength to get to the front box, so it sums in phase out front. Because sound diffracts around the rear box almost perfectly, it makes very little difference whether it is facing forwards or backwards, and you can't think of turning it around as a polarity inversion. Note that the second method doesn't work as well as the first, because the individual box does provide some degree of forward directivity, and there is slightly less (3dB or so) less sound behind the box than in front. This leads to a more hypercardiod-like pattern, which can be mitigated by turning down the rear sub about 3dB. But, by turning down the rear sub 3dB you lose that energy in the main lobe: |
Mike Pyle wrote on Fri, 28 January 2005 23:45 | ||
One thing I still am not clear about: If this array is set up like the first example you mention here, Mike B's way, is there still summation at the front of the array even though the rear sub is delayed? Or is the output of the rear sub substantially expended in cancelling the spill from the front sub? If there is summation, how much is the sound affected by the delay in the rear sub? Is my thinking right that the combination of distance, signal delay and polarity inversion will make it a full cycle off? |